07/22 — LNDN
︎ 08
all YIN
no YANG.
︎︎︎
︎ All Yin No Yang
2022
︎ Research
︎ Machine Learning
Publication:
︎
2022
︎ Research
︎ Machine Learning
Publication:
︎
Artistic proposal that seeks to build a dialogue between
contemporary machine learning methods for image generation and the process of
individuation, understood here as articulating the becoming of form.
We explore the varieties of formal divergence made possible using text-to-image diffusion models, probing the efficacy of using semantic description as a constraint to parameterise aesthetic variation within an original dataset of oil paintings.
In this work, the model takes on the active role of a critic, interpreting the images provided to it under the rubric of a text-prompt which subsequently guides the generation process, producing an environment in which in which diffusion and description can act in tandem to amplify latent potentials within an image-based practice.
We explore the varieties of formal divergence made possible using text-to-image diffusion models, probing the efficacy of using semantic description as a constraint to parameterise aesthetic variation within an original dataset of oil paintings.
In this work, the model takes on the active role of a critic, interpreting the images provided to it under the rubric of a text-prompt which subsequently guides the generation process, producing an environment in which in which diffusion and description can act in tandem to amplify latent potentials within an image-based practice.
I — HUMANOID SIMPLE
II — HUMANOID COMPLEX
︎︎︎